1. Field of the Invention
This invention relates to the process of depositing diamond, doped diamond and cubic boron nitride-diamond composite films. More specifically it relates to deposition of these films at high rates over large areas, based on Activated Reactive Evaporation (ARE) as first described in U.S. Pat. No. 3,791,852.
2. Description of the Background Art
The publications and other reference materials referred to herein to describe the background of the invention and to provide additional detail regarding its practice are hereby incorporated by reference. For convenience, the reference materials are numerically referenced and grouped in the appended bibliography.
Techniques used in recent years to deposit films of diamond-like carbon (i-C), diamond and boron nitride onto substrates have included chemical vapor deposition (CVD) and plasma assisted chemical vapor deposition (PACVD) involving plasma decomposition of hydrocarbon/boron containing gases. Ion beam assisted/enhanced deposition has also been used.
Diamond microcrystals were prepared using chemical vapor deposition and related techniques, at low pressures for the first time by Derjaguin and co-workers(1) by a chemical transport method. Subsequently Angus et al.(2) reported deposition of diamond onto natural diamond powder from methane gas at 1050.degree. C. and 0.3 torr pressure. They also proposed a qualitative model explaining the kinetics of diamond growth from the vapor phase. More recently Matsumoto et al.(3,4), have reported synthesis of diamond microcrystals by chemical vapor deposition from a mixture of methane and hydrogen gas in open flow systems. They have shown that the growth of diamond films can be enhanced if a heated tungsten filament is used in the CVD set up. Spitsyn et al.(5) in their paper have discussed the kinetics of diamond growth from CH.sub.4 +H.sub.2 gas mixtures. They have argued that atomic hydrogen plays a unique role in the growth of diamond from vapor phase.
Whitmell et al.(6) were the first to report the use of plasma decomposition techniques in the deposition of amorphous carbon-like films onto a negatively biased d.c. electrode using methane gas. However, the growth of films in their earlier experiment was thickness limited. This was believed to be due to the formation of an insulating film (i-C) on the surface of the d.c. biased electrode which after a critical thickness was reached, prevented the bombardment of the growing film with energetic ions from the plasma. Following that report, Holland (7) proposed a modification where an r.f. potential was applied to the electrode to achieve a constant film bombardment during growth. Using this technique Holland et al.(8,9) successfully deposited diamond-like carbon films on a variety of substrates. Over the years, many researchers have used similar processes (i.e. r.f. decomposition of hydrocarbon gas) to prepare diamond-like carbon films.(10,11) Similar techniques have been used to deposit BN films, where boron containing gases are used instead of hydrocarbon gases.
The remote plasma deposition technique developed by Lucovsky et al.(12) also falls under the category of a PACVD type process. In this process a mixture of reactive and inert gas is dissociated using r.f. excitation. The activated species, e.g., oxygen, from the plasma react down stream with the process gas such as SiH.sub.4 (for SiO.sub.2 deposition) to form complexes such as H.sub.3 Si-O-SiH.sub.3 in the gas phase which subsequently condense on the substrate. Bombardment by energetic neutrals dissociate the complex to produce the compound films. This technique has been successfully used by Richard et al.(13) to prepare SiO.sub.2, Si.sub.3 N.sub.4 at low deposition temperatures. They have proposed to extend this technique to the deposition of diamond by using CH.sub.4 as a process gas and H.sub.2 or a H.sub.2 +He gas mixture for activation.
Aisenberg and Chabot(14) were the first to report deposition of diamond-like carbon films by ion-beam deposition of carbon. Attempts to deposit similar films using magnetron sputtering and r.f. sputtering were only partially successful. It is likely that negligible substrate bombardment in the case of magnetron sputtering and substrate overheating in case of r.f. sputtering may have restricted the formation of i-C films in the above two techniques.
However, the dual ion beam technique used by Weissmantel(15,16) has proved to be quite successful in synthesis of diamond-like carbon films. He used a primary beam to deposit carbon with the growing film being simultaneously bombarded by Ar.sup.+ ions generated from the second ion source. Weissmantel has successfully used this technique to deposit i-C, i-BN as well as i-C/i-BN composite coatings.
In plasma decomposition techniques, the rate of deposition of the carbon films critically depends on the rate of dissociation of the hydrocarbon gas. To increase the dissociation rate, one has to increase the gas pressure and/or the r.f. power used to excite the plasma. However, the increase in r.f. power also increases the energy of the bombarding species. Moreover, increased dissociation of hydrocarbon gas produces a greater amount of hydrogen that can be trapped into the growing films--thereby producing excessive stress in the film.
A modification has been suggested where independent sources are used, one to dissociate the hydrocarbon gas, and the other for film bombardment. One such modification is due to Nyaiesh et al.(17) who have used separate r.f. sources, one to dissociate the hydrocarbon gas and the other for substrate biasing which in turn controls the bombardment of growing film. Though this technique has shown some improvement in deposition rate, the authors note that the substrate bias was affected by the power applied to the r.f. oven. Moreover, they report that input power to the r.f. oven was limited due to deposits formed by polymerization onto the chamber walls, which reduced the deposition rate.
Another approach is proposed by Kamo et al.(18), Saito et al.(19), and also by Doi et al.(20,21), where a microwave discharge is used to decompose the hydrogen gas and an independent r.f. source is used for substrate biasing. These authors have reported deposition of i-C, diamond and boron doped diamond films using this technique. However, this technique does not appear to be much different than that of Nyaiesh et al.(17) and would therefore suffer from similar limitations. In fact, the optimum deposition rate reported by Doi et al.(21) is about 1 um/hr. which seems to be very low. Moreover, even with the above-proposed modifications, it is not possible to control the hydrogen content of the films independently of the other process variables.
Although the ion beam technique provides advantages as regards independent control of substrate bombardment, deposition rate and hydrogen content, it suffers from the following two major limitations: (1) low deposition rates due to the low sputtering yield of carbon; and (2) limitations for large area deposition due to limitations in the available sizes of the ion sources.
Strel'nitskii et al.(23) have reported deposition of i-C films using energetic C.sup.+ ions from an arc source.
As is apparent from the above background, there presently is a continuing need to provide improved processes for depositing diamond and diamond-like films on substrates. Such improved processes should be able to provide control over the rate of generation of reaction vapors e.g. C, B, etc. independently of other process parameters. The process should also provide control over the plasma volume chemistry independent of the other process variables and provide control over the film bombardment independent of the other process variables. These attributes in such a process will make it possible to deposit diamond, doped diamond and cubic boron nitride-diamond films at higher rates and over large areas.